The association of cigarette smoking and alcohol drinking with body mass index: a cross-sectional, population-based study among Chinese adult male twins

Background: Obesity is a multifactorial abnormality which has an underlying genetic control but requires environmental influences to trigger. Numerous epidemiological studies have examined the roles of physical inactivity and dietary factors in obesity development. Interactions between obesity-related genes and these lifestyles have also been confirmed. However, less attention has been paid to these complex relationship between cigarette smoking, alcohol drinking and obesity. The purpose of this study was to assess whether cigarette smoking and alcohol drinking were associated with body mass index (BMI), and whether these lifestyle factors modified the genetic variance of BMI. Methods: Subjects were twins recruited through the Chinese National Twin Registry, aged 18 to 79 years, and the sample comprised 6121 complete male twin pairs. Information on height, weight, cigarette smoking and alcohol drinking status were assessed with self-report questionnaires. The associations of cigarette smoking and alcohol drinking with BMI were evaluated by linear regression models. Further, structure equation models were conducted to estimate whether cigarette smoking and alcohol drinking status modified the degree of genetic variance of BMI. Results: After adjustment for a variety of socio-demographic and lifestyle factors, former smokers had higher BMI (beta = 0.475; 95 % CI, 0.196 to 0.754) whereas moderate to heavy smokers had lower BMI (beta = -0.115; 95 % CI, -0.223 to -0.007) when compared with nonsmokers. BMI decreased with increased cigarette pack-years (beta = -0.008; 95 % CI, -0.013 to -0.003). These effects still existed substantially in within-MZ twin pair analyses. By contrast, current alcohol drinking had no significant influence on BMI when additionally controlled for shared factors in within-pair analyses. Genetic modification by alcohol drinking was statistically significant for BMI (beta = -0.137; 95 % CI, -0.215 to -0.058), with the intake of alcohol decreasing the additive genetic component of BMI. Conclusions: Cigarette smoking was negatively associated with BMI independent of genetic influences. The influence of genes on BMI was moderated by alcohol drinking, such that for individuals who were regular drinkers, genetic factors became less influential. Our findings highlight gene-alcohol interaction in finding candidate genes of BMI and elucidating the etiological factors of obesity.